India's Ministry of Electronics and Information Technology conducts competitive bidding for procurement of 10,000 GPUs from multiple vendors.

The government has selected 10 companies for the final bidding process to procure 10,000 graphics processing units (GPUs) under the IndiaAI Mission. The selected firms are Jio Platforms, Tata Communications, CMS Computers India, NxtGen Datacenter and Cloud Technologies, Ctrls Datacenters Ltd, E2E Networks, Locuz Enterprise Solutions, Orient Technologies, Vensysco Technologies Limited, and Yotta Data Services. These companies were selected after submitting applications the previous week, with the bidding process commencing on Wednesday and results expected within a week.

The India AI Mission, approved by the Cabinet in March 2024, carries a total outlay of ₹10,372 crore over five years to encourage AI development in the country. As much as 44 per cent, or ₹4,563.36 crore, of the total outlay has been earmarked specifically for providing the compute capacity of these 10,000 GPUs. The GPUs will be made available to various stakeholders including academia, MSMEs, startups, governments, and public sector agencies to support the creation of an AI ecosystem.

The procurement process underwent refinements to broaden participation. MeitY met with 13 companies individually, selected from 19 applicants seeking the opportunity, with 10 ultimately shortlisted for final bidding. Three firms—CloudThat Technologies, Hostin Services, and I2k2 Networks—did not make the final selection. Following industry feedback that the tender provisions were restrictive for smaller players, the average turnover requirement for primary bidders was lowered from ₹100 crore to ₹50 crore, while the requirement for non-primary consortium members was halved to ₹25 crore, as documented on September 27, 2024.

GPUs are vital for AI infrastructure as they can handle many tasks simultaneously, making them much faster than regular computer processors. The speed is crucial for training and running AI models, which must process large amounts of data quickly. According to leading GPU chipmaker Nvidia, India's computing infrastructure represents less than two per cent of global capacity, a limiting factor in the country's research contribution, which similarly remains at approximately two per cent. As demand for GPU-based servers and accelerated computing continues to increase due to their superior data processing speeds compared to CPU-based servers, the Centre is working to develop its own infrastructure and install GPU servers across India.